2016/12/19(月) 15:00 開催
東京都 テレコムセンター

第11回産総研人工知能セミナー「Biomedical Text Mining: impact, challenges and applications」


日 時: 2016/12/19(月) 15:00 〜 18:00
会 場: 産総研 臨海副都心センター 別館11階 会議室1
住 所: 〒135-0064 東京都江東区青海2-4-7


第11回産総研人工知能セミナー「Biomedical Text Mining: impact, challenges and applications」

産業技術総合研究所 人工知能研究センターでは、人工知能研究に関する情報交換を目的として、原則として月に一度、外部の方やセンター内研究者を講師とする人工知能セミナーを開催しています。12月は「Biomedical Text Mining: impact, challenges and applications」を開催します。



生命科学分野では、多様なデータベースが開発・構築されていますが、こうしたデータベースの基幹の一つとなるのが、膨大な文献情報です。しかしながら、こうした膨大な文献情報を解析するキュレータ人材なども不足し、効率良く文献情報を解析、抽出し、有用な知識として活用する技術として、自然言語処理(NLP)技術、テキスト・マイニングが必要不可欠となってきています。今回、英国マンチェスター大学のNaCTeMから、Sophia Ananiadou教授及びRiza Theresa Batista-navarro博士を、米国NIHのNLMからDina Demner-Fushman博士をお招きして、NLP、テキスト・マイニングの生命科学分野での応用例などについて、ご講演頂く予定です。


  • 他の方に参加の機会をお譲りするためにも、参加ができないと分かった場合は早めのキャンセルをお願いします。
  • 本名でのご登録をお願いします。
  • 産総研は、お送りいただいた情報をセミナー運営以外の目的には使用しません。
  • 講演は英語で行われます。通訳は付きません。
  • 懇親会の予定はありません。



  • Prof. Sophia Ananiadou (Manchester Univ.)
  • Dr. Riza theresa Batista-navarro (Manchester Univ.)
  • Dr. Dina Demner-Fushman (NIH)


時刻 内容
15:00-16:00 "Text Mining bridging the gap between knowledge and text"

Prof. Sophia Ananiadou (Manchester Univ.)

Text mining plays a key role in automatic semantic metadata extraction, driving the extraction of structured information from unstructured documents in the form of named entities or fine-grained and often complex relations between them (events). Event extraction techniques are used in biomedicine to extract structured representations for applications such as semantic search and pathway construction. Increasingly it becomes important not only to extract events but also to capture their contextual interpretation such as certainty, negation, source, etc. I will describe current methods of event extraction used for pathway construction and advanced semantic search.
16:00-17:00 "Development of bespoke curation workflows through the integration and federation of NLP and text mining methods"

Dr. Riza theresa Batista-navarro (Manchester Univ.)

Argo ( is a generic text mining workbench that can cater to a variety of use cases, including the semi-automatic annotation of literature. Based on the Unstructured Information Management Architecture (UIMA) standard, it enables its technical users to build their own customised text mining solutions by providing a wide array of interoperable and configurable natural language processing (NLP) components that can be seamlessly integrated into processing workflows. This talk will present the various methods which have been integrated into Argo to facilitate the development of various types of bespoke text mining workflows ranging from ones which enable the incorporation of information from various controlled vocabularies, to those which train and apply machine learning-based concept recognition models, through to user-interactive ones which support semi-automatic curation of databases, in which human annotators manually provide their corrections to automatically generated annotations. In order to allow the federation of various tools which have been developed by various groups in the NLP and text mining communities, Argo has been extended to enable interoperation with tools wrapped as web services conforming with the Representational State Transfer (REST) protocol, a widely embraced web standard. As a result, the workbench now supports the construction of NLP and text mining workflows that seamlessly incorporate even tools from other groups specialising in NLP. Specific applications of Argo's capabilities will be discussed in detail, including the curation of metabolic reactions, signalling pathway interactions, and COPD phenotypes.
17:00-18:00 "NLM's research and resources for biomedical natural language processing"

Dr. Dina Demner-Fushman (NIH)

The talk will cover clinical and consumer health question answering, clinical decision support, summarization and indexing of full text articles for retrieval, and text processing tools, services provided by the Indexing Initiative project and an upcoming FDA challenge on extraction of adverse reactions from drug labels.
  • Twitterでシェア
  • 0
  • 0
  • 0


2016/12/11(日) 12:00 〜 19:00
東京都 飯田橋
2017/01/10(火) 16:00 〜 18:00
東京都 テレコムセンター